Innovative Projects Realized

Control Forecasting Feature

Control is a leader in mobile payment analytics and alerts for SaaS, subscription, and eCommerce businesses, enabling instant intelligence anywhere via its Android, iOS, and web-based products. We collect our customers’ payment data and provide them with their key business metrics that helps them monitor their performance. In order to improve our service to customers we are moving towards providing customers with predictive analytics, that is we want to enable our users to forecast the future of their business based on the models that will learn from their historic payment data and yield forecasts for short-term future. This will help business owners to make informed business decisions and set their strategies on insights that are driven from their data. This requires studying the data and testing forecasting techniques to measure the performance of each method in predicting the future data, which is the main objective of this project.

Effectiveness of EarlyDetect, a psychiatric assessment program, using a mobile APP interface

Mental health illness cost over $50 billion in the Canadian economy annually, and is prevalent at workplace. The cost and prevalence of menial health illness can be greatly mitigated with early detection and intervention. EarlyDetect is a program developed by Chokka Center for Integrative Health for convenient, accurate and reliable self-assessment of mental illness, and has been validated in a pen-and-paper form. The current project is to develop a fully functional mobile APP version of the EarlyDetect, and to study whether testing through a mobile APP will generate the same high quality diagnosis from the pen-and-paper version of the program.

Application de l’apprentissage automatique pour l’analyse de données dans le contexte Industrie 4.0

Le Centre de recherche industriel du Québec (CRIQ) a pour mission de contribuer à la compétitivité des secteurs industriels québécois et à la croissance des organismes en soutenant l’innovation, la productivité et les exportations. Dans le cadre de ces deux sous-projets MITACS, nous allons explorer comment construire et exploiter des modèles obtenus par apprentissage automatique afin d’aider le CRIQ dans son mandat. Une partie des activités de ce projet porteront sur la recherche et l’extraction d’information à des fins de veille technologique. Le CRIQ mène fréquemment des recherches pour trouver les normes qui s’appliquent à la commercialisation d’un nouveau produit. Le premier sous-projet, qui fera l’objet de six stages de maîtrise, aura pour but d’évaluer l’apport de techniques d’apprentissage automatique et du traitement automatique de la langue naturelle (TALN) pour accomplir certaines étapes du processus de recherche d’information. TO BE CONT’D

Potentiel de l’Hyper-Cam Méthane aéroportée pour la détection des émissions surfaciques de méthane sur les lieux d’enfouissement sanitaires

Telops, une entreprise canadienne spécialisée dans la conception et la fabrication d’imageurs hyperspectraux et de caméras infrarouge, a développé l’Hyper-Cam Méthane, un outil permettant de réaliser des mesures précises des émissions surfaciques de méthane. Cet outil est particulièrement adapté à la détection et à la quantification des émissions de méthane sur les sites d’enfouissement, qui peuvent être une source importante de gaz à effet de serre. Cependant, il n’a pour le moment été utilisé que sur des supports terrestres, ce qui permet d’obtenir une quantification des émissions de méthane sur des superficies restreintes uniquement. L’objectif de cette étude est ainsi de tester les capacités de l’Hyper-Cam Méthane aéroportée pour la détection et la quantification des émissions de méthane. TO BE CONT’D

Assessing the Impact of an Electronic Patient Tracking Board onOrganizational Performance

The benefit of electronic patient management systems within hospitals has been

previously demonstrated through enhanced communication and collaboration

between departments. However many hospitals within the Canadian healthcare system

do not currently use electronic tracking systems. Furthermore, the implementation of

these systems has been challenging leading to poor adoption of these technologies

resulting in limited benefit. In order to identify the drivers and barriers to technology

adoption we intend to study the implementation of an electronic patient management

system in a large hospital setting. By interviewing and surveying those responsible for

the acquisition of the system as well as those who interact with it, the effect of this

technology on the hospital operations can be evaluated. Identification of the factors

leading to adoption will allow McKesson to design better user interfaces and

deployment strategies to increase end-user interaction and satisfaction.

Using structure-from-motion and 3D analysis to visualize habitat complexity and dynamics of glass sponge reefs in British Columbia

Glass sponges build their skeletons out of silicon dioxide (i.e. glass). While these animals are found all over the world in very deep water, they only exist shallower than 50 m in a few places in the world. In very rare cases, new sponges grow on top of existing, dead sponges and form reefs in a similar manner to coral reefs. As with coral reefs, the structure formed by the reefs is ecologically important because it provides complex habitat and shelter for other animals. However, the structure of the reef can be challenging to measure using the two-dimensional methods that ecologists traditionally use. We propose to measure the glass sponge reefs using technology that captures 3-D shapes using photos or videos. This same technology is used to make 3-D animations and virtual reality, and can allow researchers to visualize information in ways never before possible. TO BE CONT”D

Reducing homelessness using direct cash transfers and motivational interventions

This project aims to examine the effect of unconditional direct cash-transfers to recently homeless individuals in Vancouver, BC, Canada. We will conduct a randomized controlled trial to distribute a one-time unconditional $7,500 cash lump sum to 50 recently homeless individuals from shelters in Vancouver. We will also provide a supplementary motivational training for half of the participants. We will also recruit 50 recently homeless individuals as a control group, and half of these participants will also receive motivational training. At 3, 6, 9, and 12 months after the cash transfer, we will follow up with the participants to evaluate the effect of the cash transfer and motivational training. Specifically, we will measure participants’ cognitive function, social connections, housing outcomes, employment outcomes, and health outcomes.

Advancing an Artificial Intelligence Platform for Crop-Health Monitoring

Plants can respond to changes in their surroundings and can convey precise information about their health state. Ecoation has developed a multi-sensory data acquisition device to capture this information and has been collecting in-field sensor data along with data labels produced by human experts during data collection. In addition, images of various parts of plant canopy has also been collected to supplement the sensory information and to provide insights into plant physical features such as vegetation. There is a wide spectrum of data types available, but only a few have been explored due to the novelty of the data type and the application. The focus of this project will be (i) conducting research and developing new data processing methods for our data and (ii) investigating machine learning and deep learning procedures to build deployable models for plant health diagnosis, vegetation detection and other grower’s task using collected labeled data.

Analytique prédictive de l’état de santé des actifs technologiques

Le but de ce projet est de mettre en place un modèle de machine learning afin de détecter les pannes sur les actifs informatiques avant que celles-ci ne se présentent. La maintenance prédictive permet de monitorer l’état de santé des actifs informatiques et de prendre des décisions appropriées lorsque ceux-ci se dégradent. En effet une panne de serveurs peut avoir des répercussions négatives sur les clients de la Banque Nationale du Canada (distributeur automatique de billet non fonctionnel, site web inactif, panne du système transactionnel) et il faut donc autant que possible limiter ces évènements. L’étudiant, dans le cadre de son projet supervisé à la Msc. en Intelligence d’affaire à HEC Montréal devra donc développer des modèles prédictifs afin de détecter les disfonctionnements sur les différents actifs informatiques de la Banque Nationale. Une attention toute particulière sera portée sur les liens entre les pannes.

Examining Canadian Student Smoking Behaviours

The project is to examine the determinants of e-cigarette, tobacco and marijuana use among Canadian students. Specifically, we intend to determine what are the major risk factors associated with e-cigarette, tobacco and marijuana use, and to identify strategies that can be used to help address these issues. Statistical analysis will be conducted to measure the relationship between the risk factors and outcome (e-cigarette, tobacco and marijuana use). The Lung Association of Saskatchewan can use information from this study to develop easy to understand resources and messaging for the public, and results of this project can help them advocate to all three levels of government for policy and legislative change.

3D RFM SLAM for UAVs Subject to Severe Vibration

The first step for any robot to achieve true autonomy is to create a map of its surroundings and localize itself within this map at the same time. This is popularly known as the Simultaneous Localization and Mapping (SLAM) problem. Although much theory has been developed over the years to solve the SLAM problem, researchers have been having difficulties in real-world application. This is because sensors and actuators onboard a robot are always corrupted by noise. In particular, Unmanned Aerial Vehicles (UAVs) face additional difficulties that land vehicles do not. Because UAVs travel in 3D rather than 2D, additional nonlinearities associated with rotation complicate the SLAM solution. Furthermore, excessive vibration during flight can render sensitive onboard sensors useless. TO BE CONT’D

Tree to Home Scoping

This research aims to identify the challenges and opportunities that exist within and amongst First Nations communities for creating culturally appropriate and resilient housing models and local economies. It will build understanding around what is required to transform the current housing system from one that uses external labour, resources and businesses to one that promotes local livelihoods, value-added businesses and local resource use within a First Nations culturally focused bioregional economy. To accomplish this the intern will investigate financing options, policy and legislative constraints or opportunities, value-added services, and the ways in which First Nation led land use plans can be operationalized to stimulate local economic development. This research will play a key role in forwarding Ecotrusts’s mission to develop alternative approaches to economic development that can generate greater long term benefits for communities of place in terms of improved social, cultural, ecological and financial outcomes.